Move Servo IPC off UI thread

Root cause of the post-tab lag: the GPUI 16 ms timer was calling
`WebSurfaceRuntime::ensure_tab` and `tick` on the UI thread, and each
call did a synchronous `serde_json` write plus `read_line` against the
Servo sidecar over stdin/stdout. With even one visible tab, every
frame stalled on cross-process IPC.

Introduce `web_surface_worker.rs` — a per-profile worker thread that
owns the `ServoLiveClient`, drains a coalescing request queue
(latest Ensure/Poll per tab wins, no unbounded growth), and ships
results back through a `std::sync::mpsc` channel. `WebSurfaceRuntime`
now submits work non-blockingly and drains responses in `tick`; the
UI thread never blocks on the sidecar.

Adjacent in-flight cleanup riding along: hardware IOSurface
rendering-context completion (sidecar `live_protocol`,
`hardware_rendering_context`, GPUI BGRA surface shader), CSS viewport
size + device pixel ratio plumbing into `ServoLiveFrame`, and the
Send opt-ins for `CVPixelBuffer`-bearing types so frames can cross
the thread boundary.

crates/ely_app/src/services/servo_live_wire.rs

@@ -96,6 +96,12 @@ pub(super) struct LiveFrameReport {
     pub(super) state: String,
     pub(super) width: u32,
     pub(super) height: u32,
+    #[serde(default = "default_device_pixel_ratio")]
+    pub(super) device_pixel_ratio: f32,
+    #[serde(default)]
+    pub(super) css_viewport_width: u32,
+    #[serde(default)]
+    pub(super) css_viewport_height: u32,
     pub(super) rgba_byte_count: usize,
     #[cfg(all(test, feature = "live-site-smoke"))]
     pub(super) non_white_pixel_count: u64,
@@ -105,6 +111,10 @@ pub(super) struct LiveFrameReport {
     pub(super) sample_hash: u64,
 }
 
+fn default_device_pixel_ratio() -> f32 {
+    1.0
+}
+
 /// Per-frame tag that tells the renderer which already-imported
 /// `MTLTexture` to sample. Emitted at `trace` instead of `info` because
 /// it fires every frame on the hardware path; the import event above